基于CLBP、改进KPCA和RF的牛肉大理石纹评级

Beef marbling grading based on CLBP improved of KPCA and RF

为进一步提高牛肉大理石纹评级的正确率,提出了基于完整局部二值模式(Completed Local Binary Pattern,CLBP)、改进核主成分分析(Kernel Principal Component Analysis,KPCA)和随机森林(Random Forests,RF)的牛肉大理石纹评级方法。首先,利用CLBP提取牛肉大理石纹图像的纹理特征;其次,采用混沌蜂群算法对KPCA的核参数进行优化,使KPCA的降维效果和特征提取达到最优,获得表征牛肉大理石纹样本图像的特征向量;最后,使用随机森林完成牛肉大理石纹样本的分级识别,获得最终评级结果。大量实验结果表明,与基于分形维和图像特征的方法、基于灰度共生矩阵和BP(Back Propagation)神经网络法相比,本文方法所得识别率最高。

In order to further improve the correct rate of beef marbling grading, a beef marbling grading method based on completed local binary pattern（CLBP）, kernel principal component analysis and random forests is proposed. Firstly, CLBP is used to extract texture features of beef marbling image. Then the kernel parameters of kernel principal component analysis is optimized by chaos artificial bee colony, which makes dimensionality reduction and feature extraction by KPCA to achieve the optimal effect. Thus feature vectors are obtained to characterize the beef marbling images. Finally, random forests is applied to complete classification recognition of beef marble samples and get the final ratings result. A large number of experimental results show that, compared with method based on fractal dimension and image features, method based on gray level co-occurrence matrix, and the method based on Back Propagation neural network, the proposed method attains the highest recognition rate.